<<I am very curious how other folks are deriving their discounts….What say you?>>I have learned to use financial data as of the valuation date and I use the capitalized/economic income method/approach to est. equity fair market value (FMV) as commonly defined to compare to price so I want market risk premium data not my own required return. When you quote the media & boards saying “market is overpriced”, I find they give little specifics, muddled dates and mixed value approaches/methods. For me, understanding them is not possible & I pay little attention. I might have an opinion if I est. the top 10 SnP FMV vs. price. I do not give the MSM or analysts much credit for doing that very well. You ask an interesting question though. I use the build-up method also, based on market inputs (since I am going after FMV).Cost of debt<<Risk free + company specific risk>>I figure on the valuation date the company could market all new debt pretty much at the weighted avg. cost (interest expense/total debt) * (1-TR). So that is my after tax cost of debt i.e. historic, what the company is really paying out. You could adjust this a bit better to market if recent debt is selling for say 105 or 95 but I do not take the time. Business appraisers would do this. It adds some meaningful precision but time is limited.NB: A headache here is that many companies report on I/S “interest expense, net”. It is net of interest income meaning you have to go figure that out from the Notes and account for it. Some do so w/o using the “,net”, you can’t be too careful.Cost of Equity<<Risk free + company specific risk + equity risk premium>>Here I use Ibbotson’s Year Book information. Their risk free is the 20-yr UST-Bond and they compute an ERP above it based on S&P500 equity market returns since 1926 & update annually. For company specific risk, I don’t use the co. bonds since they have different terms & market dynamics. For me, I want to know the company specific equity risk vs. S&P500 equity risk during a 5 year look back period. That’s my compromise for the important to know but unknowable future specific equity risk vs. SnP equity risk during a long holding period. This limits my circle to steady eddie types of businesses. An analyst has got to know his limitations; this is one I live with. Company specifics change over longer terms and shorter terms reduce specific risk eventually to noise.Ibbotson’s has some nice work on co. size risk premia, meaning investors want bigger discounts on the riskiness of small relative to large market cap.; I can buy this point. So I add a discount premium term for size:Risk free + company specific risk + Ibbotson equity risk premium + Ibbotson size risk premiumNow I allow there is also unsystematic risk. For example, I have run into small companies with huge cash positions relative to assets. Management is scared stiff about surviving the next downturn. This is a more stable situation than your average dotBust or bioPharma. I am willing to add another unscientific guess term (but I rarely do) - this might be used to help adjust to market price if I can’t find anything else (e.g. indicates how the market values safety of enormous cash piles) thus:Risk free + company specific risk + equity risk premium + size risk + unsystematic riskover pricing… many folks are currently over pricing risk…The RFR & ERP are from the economy as a whole, whatever LT bonds & equity market prices, it is the current deal. Go into any country and that country economy sets these things w/volatility. Now you can go read a lot of gurus and get their input, I’m not going there. About all I would allow is short rates are now artificial low due to the Fed. decision but I don’t use short rates for any analysis except cash interest income & complaining about my bank account returns, so I try to minimize that issue. If you ask me for my guess at the variable of importance in market price over/under media reports, it is the next year’s est. of EPS or FCFE not discounts. The “overpricing” is also “underyielding” and that yield numerator (E or FCFE) is easier to “get” in the analysis of financial statements than the denominators (discount, growth). YMMV. KO ExampleNow let’s use it in a sentence. Derive KO discount. Valuation date 12/31/2009.Value Line has KO 5-yr specific equity risk 60% of the market ERP so it reduces ERP by 2.68% & here’s some other data as of value date: RFR = 4.58%, Co. Specific Risk Premium = -2.68%, Ibbotson ERP = 6.7%, Ibbotson Size premium = -0.37%, unsystematic risk = 0, 2009 Interest exp. = $355Mln, Total debt= $11,859Mln, TR = 2040/8946. (No implied warranty on math or numbers, double check.) Cost of Equity I’d use = 4.58 - 2.68 + 6.7 -.37 = 8.23%, rounded.Cost of Debt I’d use = 3% * (1-.228) = 2.31%, rounded.WACC = (.12 * .0231) + (.88 * .0823) = 7.5%, rounded.